A few critics of the safeguards system contend that there is a contradiction between these two functions. The most extreme form of this argument which I have met, is that safeguards inspectors of the IAEA do not want to discover any irregularities, because this would prove that the promotional work of the IAEA had contributed to misuse. Such argument has nothing to do with the practical reality. Inspectors routinely report anomalies which they find and these are fully explored. The argument also flies in the face of the original "Atoms for Peace" policy, which deliberately held out the bargain of offering peaceful nuclear techniques and obtaining commitments to non-proliferation and acceptance of verification. There is also no doubt that in many cases effective safeguards are more readily accepted because the IAEA is identified not only as an inspector but also as an organization which helps development. Although it is true that some assistance will expand the general fund of knowledge of things nuclear, the assistance we give is mostly focussed on medicine, agriculture, industry and safety or established nuclear power programmes. I suspect that those who criticize the promotional role of the Agency do it less out of genuine concern for the non-proliferation and effectiveness of safeguards than because they want to conclude that the safeguards system is weak or because they are opposed to any promotion and use of nuclear techniques. Whatever explanation, it is obvious that there is a massive support in the Agency membership for the dual role which the Statute lays upon the IAEA and an insistence by developing countries that there continue to be a balance between the two major activities. In this view, an expansion of safeguards activities should be complemented by an expansion in promotional activities.

The legal basis for safeguards

To understand how safeguards function we must understand how the international community functions. Rules binding States do not arise in the same way as rules arise within States and the means through which rules are upheld vis-à-vis States -- when they are upheld -- are different from the means by which rules are upheld within States. Within a State a legislature can adopt by majority vote a law under which, say, citizens are allowed to possess weapons only if they have a license. The police will possess the means to enforce this rule in all cases which come to their knowledge.

In the international community there is no legislature authorized to decide by majority that only the five declared nuclear-weapon States are to be licensed the possession of nuclear weapons and although references are not rare to "a norm of non-proliferation", this is a "political" norm -- not a legally binding one. Clearly, however, it is a norm with considerable backing. In their summit statement of 31 January 1992 the members of the Security Council stated inter alia that "the proliferation of all weapons of mass destruction constitutes a threat to international peace and security" -- thereby implying a capacity for the Council to intervene. However, any attempt to maintain that a non-proliferation norm exists today which legally excludes the nuclear weapons option for all non-declared nuclear-weapon States would I think be opposed by Israel, India and Pakistan. The development of legally binding commitments to non-proliferation in the international community is not through majority votes but by the more cumbersome process of individual express consent of States to various treaties. This process as we know has been remarkably successful and now misses only a few crucial countries -- apart from the five declared nuclear-weapon States. We have also seen in the DPRK case how a State's notice to revoke its consent led to immediate and vigorous reactions. It is safe to assume that the NPT Conference in 1995 will not lead to a lapsing but to a prolongation of the legally binding non-proliferation commitments made under the Treaty.

It is correct that from the viewpoint of any State, say on the Arabian peninsula, it would appear desirable that the Agency operate as fully effective comprehensive safeguards in Israel as in Iraq. And we are sometimes criticized for not doing this. However, it must be understood that, unlike Iraq, Israel has not yet adhered to the NPT and has not yet accepted a comprehensive safeguards agreement. Therefore the Agency has no right to implement such safeguards in Israel. This is not a weakness of the safeguards system, but a consequence of state sovereignty. The first step is to seek the contractual consent of Israel. In the same way, the IAEA cannot be criticized for not applying comprehensive safeguards in India or Pakistan, which have also not adhered to the NPT.

Is it justified to criticize the safeguards system as weak, because the inspectors cannot forcibly obtain access to countries and installations? We must realize that enforcing rules on States is mostly a very different matter from enforcing rules on individuals in a State. The inspectors of the IAEA are not an international nuclear police that can force its way to facilities it feels it must visit. Indeed, a country can even make the inspectors’ entry subject to visa. To exert pressure the Agency can suspend technical assistance or even suspend rights and privileges of a country, but such actions do not constitute very effective sanctions. However, the Agency can report a case of non-compliance to the Security Council, which possesses a wide spectrum of legal competences to bring a State to fulfil its obligations. It is for the Council to determine in the light of all circumstances what means it finds appropriate. It may safely be assumed that military force will only rarely be resorted to. Diplomatic pressures, trade restrictions, the stopping of air traffic, economic sanctions are more likely measures. Is this to admit the impotence of the world community? Yes and no. After the Cold War, the Security Council is much more likely than before to be able to agree on some measure of persuasion or enforcement. Yet, mobilizing and applying the force necessary to coerce a State will remain problematic and will only occur where enforcement appears vitally important to the great powers in the Council. In the face of Iraqi aggression, such force was mobilized and used. In the case of the DPRK's refusal of adequate access to safeguards, great power diplomacy is still being used.

From the IAEA horizon, the DPRK case is straightforward: safeguards findings suggest that the DPRK has more plutonium than it has declared and when the country refuses to co-operate by allowing requested inspections of two non-declared sites and fails to give satisfactory explanations, the Agency turns to the Security Council. It is not really the effectiveness of safeguards as an alarm mechanism which is now being tested. Thanks to the use of new analytical techniques which are part of the strengthened safeguards, the system has produced warning signals. What is being tested are the means available to the Security Council and its Members of prompting the DPRK to demonstrate credibly through full acceptance of effective safeguards its assertion that it is not using any nuclear installations for the development of a nuclear weapons capacity. As I see it, the United States' talks with the DPRK representatives have been in pursuance of this Security Council objective.

Incentives and disincentives to nuclear proliferation

Inspection cannot -- as I said -- physically prevent or stop violation of a non-proliferation commitment, but the risk that verification could lead to the detection of violations and consequent reactions by other States, should provide some disincentive against violations.

Regrettably, despite the strengthening of safeguards, the case of a non-discovered clandestine nuclear weapons capacity cannot be ruled out. As Mr. Spector was reported to have said, "We don't know what we don't know." Accordingly we have to conclude that neither the effectiveness of the legal commitments plus safeguards and non-military incentives nor military means will give 100% assurance. However, the ambition is to come as close as we can to such complete assurance.

Before leaving this point let me finish by recalling -- even with Iraq and DPRK before our eyes -- that the policies leading to legally binding commitments to non-proliferation and safeguards have achieved a great deal, most recently regarding Argentina, Brazil and South Africa, and perhaps soon regarding Algeria, Cuba and Ukraine. The ending of the Cold War, the drastic reductions in the nuclear arsenals and the patient policies of military détente and incentives and disincentives to non-proliferation are paying good dividends.

Limits to willingness of States to accept verification and inspection

As the purpose of inspection is to justify confidence between States, it is an activity in the direct interest of neighbouring States and the world at large. Generosity concerning inspection should thus be expected. However, we find that while States may wish to see intrusive inspection in other States' territory, it may wish to see some limitations when it comes to its own territory and installations. The most recent experience of governmental discussion of this matter was in the long negotiations which resulted in the Chemical Weapons Convention. While on some points the inspection system accepted there contains clear improvements over what we have under the NPT -- e.g. regarding multiple visas and the right to communication and travel -- many features are very similar in the two verification systems, e.g. the need for acceptance of inspectors designated. The limitations are perhaps not so surprising. As citizens we do not want the police to search homes without some warrant issued for adequate reasons. It should not be expected that sovereign States will be less jealous about their privacy.

International inspection of facilities under military control may be both particularly important and particularly sensitive. The CWC solves the problem by so-called "managed access" which will allow the shielding of certain sensitive objects not relevant to chemical weapons. The problem was encountered in practice by the IAEA when the DPRK asserted that two non-declared sites which the IAEA wished to visit and take samples from were non-nuclear and military. One of them, to which the Agency was allowed access on one occasion appeared, indeed, to have some military function. In requesting access to these sites, the Agency has informed the DPRK that it is ready to discuss special arrangements which would take into account concerns about military secrecy. This is probably as far as one can go. While a State, like the DPRK, could have a genuine fear that its accepting inspection of military sites could be misused, the immunity of any sites could equally be used by a State wishing to hide prohibited material or installations.

In my view, objections to the use of satellite photography or intelligence data by an inspectorate deserve less understanding. Satellite pictures and remote sensing have been part of confidence-building for a very long time and must by now be regarded as conventional surveillance. It might be good for an international inspectorate to obtain pictures from several different systems -- to reduce the risk of any manipulation. National intelligence offered to an international inspectorate may, of course, be disinformation or erroneous. So long as the inspectorate is fully aware of this possibility and critically examines any such data, I do not see strong reasons for objecting to its use -- if it is offered. There will undoubtedly be restrictions in States’ readiness to offer such data. However, a State can hardly complain about ineffectiveness of an international verification system if it silently sits on information that could have helped the international system to examine particular sites. Both the Special Commission of the UN on Iraq and the IAEA received valuable information from national intelligence during their inspection work in Iraq. It is, on the other hand, excluded, of course, that an international inspectorate could engage in illicit activities, or offer any State information in return for intelligence obtained.

The conclusion that 100% certainty is unattainable through safeguards verification takes on greater significance as nuclear disarmament proceeds. The fewer nuclear weapons there remain in the world, the more important it becomes that no clandestine weapons capacity develops anywhere. No doubt this lack of full certainty will be one factor suggesting retention of a nuclear military capacity in the existing nuclear-weapon States -- or in some distant international security organ to deal with any surprises. It also points to the importance of regional and global policies which not only create but maintain disincentives to the acquisition of nuclear weapons and which maintain incentives to do without them, as well as export restrictions making it more difficult for anyone to acquire sensitive material or technology.

Alarmist reports which we read from time to time about new threats of nuclear proliferation, while useful as reminders of risks, often overlook the very significant progress that has been achieved in the realm of non-proliferation in the last few years. With Argentina having accepted full-scope safeguards and Brazil near this decision, and Cuba apparently ready to do the same, the Tlatelolco Treaty may soon come into force. And with South Africa having joined the NPT and Algeria declaring its intention to do so, an African nuclear weapons-free zone -- a draft of which already exists -- seems a very likely early project. There remain three problem areas: the Middle East, India/Pakistan and the Korean peninsula. It seems clear that regional tailor-made solutions will be needed in all three cases.

The conclusion of a comprehensive test ban treaty -- CTBT -- is of undoubted value to prevent the development of reliable nuclear weapons technology -- at least in some States. It is also of signal political importance for the continued broad support of the NPT that the nuclear weapons States join such a treaty, which has been held out as an expectation for several decades. Perhaps the time might even be ripe for a non-first use agreement. Paul Nitze has recently argued that smart conventional weapons might be more credible than nuclear weapons. A non-first use agreement -- by definition -- does not rule out nuclear retaliation.